The stereospecific Barbier reaction between (Z)- and (E)-1-iodopentafluoropropenes, zinc and aldehydes: Mechanistic aspects and scope of the preparation of (Z)- and (E)-partially fluorinated allylic alcohols

Article abstract of DOI:10.1016/S0022-1139(01)00460-2

The Zn/Barbier reaction of acid-washed zinc or zinc/silver couple with (Z)- or (E)-CF₃CF=CFI and aromatic aldehydes in DMF stereospecifically gives the (Z)- or (E)-allylic alcohol, CF₃CF=CFCH(OH)C₆H₅. Substitute

Full text of DOI:10.1016/S0022-1139(01)00460-2

Journal of Fluorine Chemistry 111 ꢀ2001) 233±240
The stereospeci®c Barbier reaction between ꢀZ)- and
ꢀE)-1-iodopenta¯uoropropenes, zinc and aldehydes
Mechanistic aspects and scope of the preparation of ꢀZ)- and
ꢀE)-partially ¯uorinated allylic alcohols
Vinod Jairaj, Donald J. Burton*
Department of Chemistry, University of Iowa, Iowa City, IA 52242, USA
Received 20 April 2001
Abstract
The Zn/Barbier reaction of acid-washed zinc or zinc/silver couple with ꢀZ)- or ꢀE)-CF3CF=CFI and aromatic aldehydes in DMF
stereospeci®cally gives the ꢀZ)- or ꢀE)-allylic alcohol, CF3CF=CFCHꢀOH)C6H5. Substituted aromatic aldehydes, which contain an electron-
withdrawing group, undergo Barbier addition reaction. Substituted aromatic aldehydes, which contain an electron-releasing group, fail to
undergo Barbier addition reaction. Aliphatic aldehydes, such as butanal, which contain an a-H, fail to undergo Barbier addition and
signi®cant amounts of CF3CF=CFH are formed. When the addition reaction is carried out in the presence of 1,4-dinitrobenzene, no reaction
is observed. Mechanistically, the Zn/Barbier addition of CF3CF=CFI to aromatic aldehydes is best explained by formation of the
[CF3CF=CF]É anion which adds to the aldehyde and is not adequately explained by formation and addition of an unsolvated zinc reagent,
CF3CF=CFZnI ꢀunsolvated). # 2001 Elsevier Science B.V. All rights reserved.
Keywords: Barbier reaction; Vinylzinc reagents; Fluorinated allylic alcohols; Stereospeci®c reactions
1. Introduction
variety of metals with the cheaper and readily available
CF3Br under pressure [6]. Kitazume and Ishikawa demon-
strated that the Zn/Barbier reaction of CF3I with carbonyl
compounds could be accomplished with ultrasound activa-
tion [7]. No reaction was observed in the absence of ultra-
sound activation. Kitazume and Ikeya have shown that zinc
and methyl viologen function ef®ciently as an electron
mediator in Barbier-type conversion of per¯uoroalkyl
iodides to per¯uoroalkyl carbinols [8] and Fujita and
Hiyama have more recently shown that Zn/Barbier reactions
with ¯uorinated organohalides are effective even without
sonication [9]. Per¯uoroiodoalkanes and per¯uorobro-
moalkanes also have been demonstrated to react with car-
bonyl compounds in the presence of zinc and a catalytic
amount of palladium or nickel reagents [10]. Wakselman
and co-workers extended their earlier work on per¯uorinated
sulfonic acids to carboxylic acid derivatives [11,12]. Recent
reports included the Zn/Barbier reaction of iodoper¯uor-
oalkanes and CFCs with aldehydes [13] and extension of this
methodology to the per¯uoroalkylation of dideoxy sugar
aldehydes [14].
Barbier reaction is a powerful tool in the arsenal of
the synthetic organic chemist for the construction of car-
bon±carbon bonds. This reaction was ®rst discovered
by Philippe Barbier in 1899. The advantage of this one-
step reaction was that the synthetic chemist could by-pass
the preparation and storage of unstable organometallic
reagents [1±4].
Previous reports of Zn/Barbier reactions involved the
reaction of iodoper¯uoroalkanes with carbon dioxide or
sulfur dioxide in the presence of Zn/Cu couple to give either
per¯uoroalkyl carboxylic acids or sulfonic acids [5].
Mechanistically, the proposed pathway was suggested to
involve insertion of CO2 or SO2 into the carbon±zinc bond of
the per¯uoroalkylzinc reagent. Under these conditions, the
stable per¯uoropropyl zinc reagent failed to react [5]. Wak-
selman accomplished a similar reaction with SO2 and a
* Corresponding author. Tel.: ‡1-319-335-1363; fax: ‡1-319-335-1270.
E-mailaddress : donald-burton@uiowa.edu ꢀD.J. Burton).
Although many Barbier reactions of carbonyl compounds
with bromo- or iodoper¯uoroalkanes have been reported, to
0022-1139/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 1 1 3 9 ꢀ 0 1 ) 0 0 4 6 0 - 2

234
V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
the best of our knowledge, only one example has been
reported for a partially ¯uorinated vinyl halide [15], and
there are no reports with a haloper¯uoroole®n.
as outlined below.
These authors demonstrated that the preformed zinc
reagent ꢀsolvated) did not add to the aldehyde. Conse-
quently, to account for the observed alcohol product,
they suggested that the zinc reagent ꢀunsolvated) was
more reactive than the solvated zinc reagent and hence
added to the aldehyde. The unsolvated zinc reagent was
not observed or generated; consequently it is dif®cult
to know the validity of this intermediate in this reaction.
When Klabunde et al. generated the unsolvated tri¯uor-
omethylzinc reagent, he showed that it reacted with
acetone or pivalaldehyde to produce ¯uoroform [18],
suggesting that the ꢀCF3 radical is involved in the
reaction.
2. Results and discussion
Mechanistically, Barbier reaction is an interesting reac-
tion and several different mechanisms have been proposed
to account for the observed results. Blomberg has provided
some fascinating insights into the mechanism of magne-
sium-promoted Barbier reactions [16,17].
In the zinc-promoted reaction of CF3Br with SO2,
Wakselman has proposed radical anion intermediates [6],
as outlined below.
M
À
‡
‡
RFI!M ‡ ‰RFXŠ ! ꢀRF ‡ XÀ ‡ M
ꢀRF ‡ SO2 ! RFSO2 ! products
À
À
ꢁCH3†3CCHO
CF3ZnIÀÀÀÀÀÀÀ! CF3H
unsolvated
In the Zn/Barbier reactions catalyzed by PdꢀII) or NiꢀII),
Ishikawa and co-workers proposed that the role of zinc is
two-fold: ꢀ1) the zinc assists the scavenging of chloride from
the catalyst to provide the zero valent metal; ꢀ2) zinc
promotes the single electron transfer reaction which pro-
vides the per¯uoroalkyl radical, which can behave in three
ways: ꢀa) it can abstract a hydrogen atom to give RFH; ꢀb) it
can self-couple to give RF±RF; and ꢀc) it can react with the
aldehyde to give the zinc alkoxide. These processes are
outlined below.
Pivalaldehyde is a poor proton donor, but an excellent
source of a hydrogen atom. On the other hand, when bro-
moform, which is an excellent proton donor but a poor
hydrogen atom source, was reacted with the unsolvated
tri¯uoromethylzinc reagent, little CF3H was formed.
In the one case involving a ¯uorine-containing vinyl
halide, namely, CF3CBr=CH2, Hu et al. also proposed an
``unsolvated'' zinc reagent to account for the zinc-promoted
addition of this reagent to aldehydes [15], and suggested that
the ``solvated'' zinc reagent ꢀpregenerated) did not add to
aldehydes. However, CuꢀI)Cl was proposed as a co-catalyst
in this reaction but does not appear in any of the mechanistic
proposals.
MꢁPPh3†2Cl2 ‡ Zn ! Mꢁ0†ꢁPPh3† ‡ ZnCl2
Zn
2
RFI ‡ Mꢁ0† ! RFMI!ꢀRF ‡ Zn ‡ IÀ ‡ Mꢁ0†
‡
ꢀ
2RF ! RF À RF
Since our discovery of the facile formation of per-
¯uorovinylzinc reagents via the direct reaction of Znꢀ0)
with per¯uorovinyliodides or bromides [19], we have had a
continuous interest in the mechanism of the zinc reagent
formation. When ``wet'' solvents are employed in the
generation of the per¯uorovinylzinc reagents, the amount
of the 1-hydroper¯uoroole®n increases with the amount
of water present. The formation of the reduced alkene
suggested the involvement of a per¯uorovinyl anion
intermediate, which captures a proton to form the 1-hydro-
per¯uoroole®n. Consequently, our ``working'' mechanism
for the formation of the per¯uorovinylzinc is outlined
below ꢀusing ꢀZ)-1-iodoper¯uoropropene as a model
substrate).
H-abstraction
ꢀRFÀÀÀÀÀÀÀ! RFH
In the reaction of RFI with aldehydes in DMF in the
presence of zinc, Shen and Qi proposed that the ®rst step of
this Zn/Barbier reaction is a single electron transfer from
‡ À
zinc to RFI [13]. Loss of Zn I results in the formation of the
‡ À
per¯uoroalkyl radical, which reacts with Zn I to form an
``unsolvated per¯uoroalkylzinc reagent'', which can react
with the aldehyde to form the alcohol ꢀafter hydrolysis),

V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
235
Since the per¯uorovinylzinc reagents did not react with
aldehydes [20], we decided to explore the Zn/Barbier reac-
tion with these zinc reagents in the presence of aldehydes to:
ꢀ1) ascertain if ¯uorinated allylic alcohols could be stereo-
speci®cally prepared by this route; and ꢀ2) probe the
mechanism of the Zn/Barbier reaction, if alcohols were
produced. Our emphasis in this investigation has focused
on the mechanistic aspects; consequently we have not
attempted to maximize the yields of the allylic alcohols
or try to remove all unreacted aldehyde from the allylic
alcohol product. Our purpose has been to demonstrate
stereospeci®city and to examine the scope of aldehydes that
would participate in the reaction.
When ꢀZ)-CF3CF=CFI was reacted in dry DMF with
benzaldehyde and Zn/Ag couple or acid-washed zinc, an
exothermic reaction ensued after a short indication period.
After hydrolysis with dilute HCl, extraction and chromato-
graphy, the corresponding ꢀE)-allylic alcohol ꢀ1) was isolated
in 54% yield. The spectroscopic data for this compound is in
agreement with previous reports of this compound [21,22].
No ꢀZ)-isomer was detected by 19F NMR.
2-fluorobenzaldehyde and zinc in dry DMF under similar
conditions, ꢂ50% yield of the ꢀZ)-allylic alcohol was iso-
lated. No ꢀE)-allylic alcohol was detected by 19F NMR. All
1
new allylic alcohols were completely characterized by H,
19F and 13C NMR and HRMS. The ꢀE)- or ꢀZ)-assignment
3
was made based on JFFꢁtrans† @ 3JFFꢁcis†. Thus, the ꢀZ)- or
ꢀE)-stereochemistry is preserved in the addition to the
aldehyde. When the aryl ring contains an electron-releasing
group, such as 4-OMe and 4-Me, the Zn/Barbier addition
of ꢀZ)-CF3CF=CFI to the aldehyde failed. No alcohol pro-
duct was detected and only ꢀZ)-CF3CF=CFZnI was formed.
The above reactions are consistent with the formation of
the ꢀZ)-CF3CF=CFÉ anion. When the electrophilicity of
the aldehyde is high ꢀelectron-withdrawing group on the
aryl ring), addition to the aldehyde is favored over trapping
‡
of the anion with [ZnI] . When the electrophilicity of
the aldehyde is decreased ꢀelectron-releasing group on the
‡
aryl ring), the trapping of the anion by [ZnI] is favored
and only ꢀZ)-zinc reagent is observed. Of course, these
results could also be rationalized by invoking an ``unsol-
vated'' zinc reagent which is sufficiently nucleophilic to
react with aryl aldehydes of high electrophilicity, but not
sufficiently nucleophilic to react with aryl aldehydes of
low electrophilicity.
Thus, the ꢀZ)-vinyliodide stereospecifically adds to the
aromatic aldehyde. We then examined a series of aromatic
aldehydes in this Zn/Barbier reaction and Table 1 sum-
marizes these results. When the aromatic aldehyde con-
tained an electron-withdrawing group on the aryl ring,
such as 2-F, 3-F, 4-F, 4-CF3, 2-CF3, 4-Cl, the Zn/Barbier
addition gives modest yields of the corresponding ꢀE)-
allylic alcohols. When ꢀE)-CF3CF=CFI was reacted with
However, when ꢀZ)-CF3CF=CFI, Zn/Ag couple, and butanal
are reacted in dry DMF, no ꢀZ)-allylic alcohol is detected,
and ꢀE)-CF3CF=CFH and ꢀZ)-CF3CF=CFZnI are the only

236
V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
Table 1
Stereospecific preparation of ꢀZ)- and ꢀE)-fluorinated allylic alcoholsa
Compound no.
Z/E ꢀolefin)
X
ꢀ%) Yield ꢀalcohol)
Z/E ꢀalcohol)
1
2
Z
Z
Z
Z
Z
Z
Z
Z
Z
Z
E
Z
Z
Z
H
54
51
51b
55
52c
48d
53e
0
E
E
E
E
E
E
E
±
2-F
4-F
3
4
4-CF3
2-CF3
3-F
5
6
7
4-Cl
4-OMe
4-Me
8
9
0
±
f
10
11
12
1
±
2-F
0
±
50g
0i
Z
±
h
±
2-F
0j
±
1
4-NO2
0
±
ꢁ1† Zn=DMF
a
ꢁZ†- or ꢁE†-CF3CˆCFI ‡ X-C6H4CHOÀÀÀÀÀÀ! ꢁZ†- or ꢁE†-CF3CFˆCFCHꢁOH†C6H4-X
ꢁ2† HCl
b 54:46 mixture with aldehyde.
c 60:40 mixture with aldehyde.
d 70:30 mixture with aldehyde.
e 74:26 mixture with aldehyde.
f ꢀCH3)3CCHO used as the aldehyde.
g 80:20 mixture with the aldehyde.
h CH3CH2CH2CHO used as aldehyde.
i ꢀE)-CF3CF=CFH and ꢀZ)-CF3CF=CFZnI only products.
j 1,4-Dinitrobenzene additive.
products detected by NMR.
This observation is consistent only with the vinyl anion
intermediate. Shen and Qi have demonstrated that aliphatic
aldehydes give high yields of alcohols in a Zn/Barbier
reaction with RFI [13]. Consequently, we would expect
allylic alcohol formation when ꢀZ)-CF3CF=CFZnI is reacted
with butanal under Zn/Barbier conditions. The formation of
large amounts of ꢀE)-CF3CF=CFH is best explained by
trapping of the vinyl anion by the enol form of butanal.
out the reaction of ꢀZ)-CF3CF=CFI, Zn, 2-¯uorobenzalde-
hyde, and 1,4-dinitrobenzene in dry DMF. No reaction
occurred either at RT or elevated temperatures. No allylic
alcohol was formed ꢀin contrast to ꢂ50% allylic alcohol in
the absence of 1,4-dinitrobenzene) and no ꢀZ)-
CF3CF=CFZnI was formed, consistent with the known
ability of 1,4-dinitrobenzene to quench SET reactions. This
inhibition of the SET process also explains why 4-nitro-
benzaldehyde failed to react with ꢀZ)CF3CF=CFI and Zn in
dry DMF ꢀcf. Table 1).
3. Conclusions
With a highly hindered aldehyde, such as pivalaldehyde, no
allylic alcohol was observed; only ꢀZ)-CF3CF=CFZnI was
detected, suggesting that the anion reacts significantly faster
with [ZnI] than with the hindered pivalaldehyde.
In order to gain some information for the formation of the
initial vinyl radical in our mechanistic scheme, we carried
The Zn/Barbier reaction of acid-washed zinc or zinc±
silver couple with ꢀZ)- or ꢀE)-1-iodoper¯uoropropenes
and aromatic aldehydes to stereospeci®cally form ꢀZ)- or
ꢀE)-¯uorinated allylic alcohols is best explained by the
formation of the ꢀZ)- or ꢀE)-CF3CF=CFÉ anion which adds
to the aldehyde. Electron-withdrawing groups on the aryl
‡

V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
237
ring promote the addition reaction, whereas electron-releas-
ing groups lower the electrophilicity of the aldehyde and
inhibit the addition reaction relative to the formation of
the vinylzinc reagent. Aliphatic aldehydes, which contain an
a-H, cause reduction of the vinyl halide. The involvement of
an ``unsolvated'' vinylzinc reagent intermediate in this
reaction is less attractive and does not adequately explain
the formation of the reduction product in reactions with
aliphatic aldehydes.
ꢀZ)-CF3CF=CFI were prepared by the published procedure
[23].
4.3. A representative procedure for Barbier preparation
of allylic alcohols from +Z)-CF3CF=CFI
4.3.1. 1-Phenyl-+2E)-2,3,4,4,4-pentafluorobuten-1-ol
+1), +E)-CF3CF=CFCH+OH)C6H5
A dry 25 ml ¯ask equipped with a side arm, stir bar and a
condenser attached to a N2 tee was charged with 1.1 g
ꢀ17 mmol) of Zn/Ag couple, 5.0 ml of dry DMF and
0.53 g ꢀ5.0 mmol) of benzaldehyde. The reaction mixture
was stirred at room temperature for 5 min. To this slurry was
added 2.59 g ꢀ10.0 mmol) of ꢀZ)-CF3CF=CFI. An exotherm
ꢀafter a short induction period) indicated initiation of the
reaction. The reaction mixture was allowed to cool to room
temperature and then stirred at RT for 2 h. The reaction
mixture was poured into dilute HCl and extracted with ether
ꢀ3 Â 10 ml). The ether extracts were combined and washed
with a saturated solution of aqueous NaHSO3 ꢀ3 Â 10 ml) to
remove any unreacted benzaldehyde. The organic layer was
then washed with brine ꢀ3 Â 10 ml), dried over anhydrous
MgSO4, concentrated by rotary evaporation, and puri®ed
by silica gel chromatography ꢀethyl acetate:hexane ˆ 5:1)
to give 0.65 g ꢀ54%) of ꢀE)-CF3CF=CFCHꢀOH)C6H5
ꢀ95% purity). 1H NMR ꢀCDCl3): 7.0±7.3 ꢀm, 5H), 5.7
4. Experimental
4.1. General
All reactions were monitored by 19F NMR analysis of
the reaction mixture on a Bruker AC-300 MHz spectro-
meter. The 1H NMR spectra of ®nal products were obtained
on a Bruker AC-300 MHz ꢀ300.17 MHz), a Bruker WM-
360 MHz ꢀ360.14 MHz) or a Bruker DRX-400 MHz
ꢀ400.13 MHz) spectrometer. Chemical shifts are reported
in ppm with respect to either internal TMS or the lock
solvent, CDCl3. 19F NMR spectra were obtained on a Bruker
AC-300 MHz ꢀ282.41 MHz) spectrometer. Chemical shifts
are reported in ppm relative to internal CFCl3 or C6H5CF3.
In some cases, the 19F NMR spectra were broadband
decoupled from protons. All 13C NMR spectra were
obtained on either a Bruker AC-300 MHz ꢀ75.49 MHz), a
Bruker WM-360 MHz ꢀ90.57 MHz), or a Bruker DRX-
400 MHz ꢀ100.62 MHz) spectrometer with CDCl3 as the
lock solvent. All 13C spectra were broadband decoupled
from hydrogen, but not from ¯uorine. Chemical shifts
are reported relative to internal TMS or CDCl3 ꢀlock sol-
vent). Low resolution mass spectral analyses were
performed at 70 eV in the electron impact mode on a
single quadruple instrument interfaced to a gas chromato-
graph ®tted with a DB-1 column ꢀ0.25 mm i:d: Â 15 m).
High resolution mass spectral analyses were performed
on a VG-ZAB instrument by the University of Iowa High
Resolution Mass Spectroscopy Facility or the Washington
University ꢀSt. Louis, MO) High Resolution Mass
Spectroscopy Laboratory in the electron impact mode.
All isolations utilizing column chromatography were
carried out on columns >12 in. in length packed with
silica gel ꢀ325 mesh). All fractions were monitored by
TLC using UV light for visualization.
ꢀdd, 4JFH ˆ4:3, 3JFH ˆ 24:5 Hz, 1H), 2.5 ꢀbs, 1H) ppm. 19
F
NMR ꢀCDCl3): À68.4 ꢀdd, 4JFF ˆ 21:8, 3JFF ˆ 10:8 Hz, 3F),
À155.5ꢀddq,3JFF ˆ135:2,3JFH ˆ 24:2,4JFF ˆ 21:6 Hz,1F),
À172.3 ꢀdqd, 3JFF ˆ 133:6, 3JFF ˆ 10:7, 4JFH ˆ4:4 Hz, 1F)
4
ppm. {1H} 19F NMR ꢀCDCl3): À68.8 ꢀdd, JFF ˆ 21:8,
3
4
3JFF ˆ 10:7 Hz, 3F) À155.5 ꢀdq, JFF ˆ 134:1, JFF
ˆ
3
3
21:6 Hz, 1F) À172.5 ꢀdq, JFF ˆ 133:8, JFF ˆ 11:1 Hz,
1F) ppm. 13C NMR ꢀCDCl3): 140.7±127.7 ꢀm), 125.5±
2
130.1 ꢀm), 63.2 ꢀd, JCF ˆ 21:6 Hz) ppm. GC±MS:
‡
238 ꢀM , 31). The 19F NMR data are consistent with
literature data [21,22].
A similar Barbier reaction using acid-washed zinc gave
54% ꢀ1).
4.3.2. 1-+2-Fluorophenyl)-+2E)-2,3,4,4,4-pentfluorobuten-
1-ol+ 2) +E)-CF3CF=CFCH+OH)-2-FC6H4
Similarly, the reaction of 0.97 g ꢀ15.0 mmol) of Zn/Ag
couple, 5.0 ml dry DMF, 0.62 g ꢀ5.0 mmol) 2-¯uorobenzal-
dehyde and 2.64 g ꢀ10.2 mmol) of ꢀZ)-CF3CF=CFI gave
0.65 g, 51% of ꢀE)-CF3CF=CFCHꢀOH)-2-FC6H4 ꢀ2)
ꢀ95% purity). 1H NMR ꢀCDCl3): 3.9 ꢀbs, 1H), 6.1
ꢀdd, 3JFH ˆ 24:9, 4JFH ˆ 3:9 Hz, 1H), 7.1±7.7 ꢀm, 4H) ppm.
19F NMR ꢀCDCl3): À75.1 ꢀdd, 4JFF ˆ 21:6, 3JFF ˆ 12:4 Hz,
4.2. Materials
5
DMF was distilled under reduced pressure from CaH2.
Zinc was activated as described in previous work [23].
Zinc±silver couple was prepared by the method of Conia
and Rousseau [24]. Bromotri¯uoroethylene and hexa¯uor-
opropene were obtained from commercial sources and
used directly. ꢀZ)- and ꢀE)-CF3CF=CFI were prepared
by the published procedures [25,26]. The zinc reagent of
3F), À125.2 ꢀd, JFF ˆ 4:2 Hz, 1F), À162.2 ꢀdm,
3
3JFF ˆ 133:2 Hz, 1F), À177.7 ꢀdm, JFF ˆ 133:3 Hz, 1F)
2
3
ppm. 13C NMR ꢀCDCl3): 62.0 ꢀdd, JCF ˆ 21:6, JCF
ˆ
4:5 Hz), 115.6±130.8 ꢀm), 110.1±154.7 ꢀm). GC±MS: 256
ꢀM , 13), 239 ꢀ10), 236 ꢀ46), 125 ꢀ76), 123 ꢀ65), 97 ꢀ100), 91
‡
ꢀ17). HRMS: calcd. for C10H6F6O: 256.0323; found:
256.0324.

238
V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
4.3.3. 1-+4-Fluorophenyl)-+2E)-2,3,4,4,4-pentafluorobuten-
1-ol+ 3), +E)-CF3CF=CFCH+OH)-4-FC6H4
Similarly, 0.97 g ꢀ15 mmol) of acid-washed zinc, 5.0 ml
dry DMF, 0.63 g ꢀ5.1 mmol) 4-¯uorobenzaldehyde, and
2.54 g ꢀ9.84 mmol) of ꢀZ)-CF3CF=CFI gave 0.67 g ꢀ51%)
of a 54:46 mixture of ꢀ3) and 4-¯uorobenzaldehyde. 1H
6JFF ˆ 3:4 Hz, 1F) ppm. {1H} 19F NMR ꢀCDCl3): À59.7
ꢀs, 3F), À68.9 ꢀdd, 4JFF ˆ 21:6, 3JFF ˆ 10:7 Hz, 3F), À155.1
3
4
ꢀdq, JFF ˆ 133:8, JFF ˆ 22:2 Hz, 1F), À171.9 ꢀdqq,
3JFF ˆ 134:0, JFF ˆ 11:0, JFF ˆ 3:6 Hz, 1F) ppm. 13C
3 6
2
3
NMR ꢀCDCl3): 63.0 ꢀdd, JCF ˆ 22:4, JCF ˆ 2:2 Hz),
124.0±134.0 ꢀm), 110.0±160.0 ꢀm). GC±MS: 306
ꢀM , 29), 288 ꢀ10), 286 ꢀ40), 161 ꢀ60), 145 ꢀ50), 77 ꢀ27),
‡
3
4
NMR ꢀCDCl3): 4.6 brs, 1H, 5.8 ꢀdd, JFH ˆ 24:8, JFH
ˆ
4:4 Hz, 1H), 7.4±8.0 ꢀm, 4H) ppm. 19F NMR ꢀCDCl3):
69 ꢀ52). HRMS: calcd. for C11H6F8O: 306.0291; found:
306.0290.
A similar Barbier reaction using acid-washed zinc gave
52% of a similar product mixture.
4
3
À68.6 ꢀdd, JFF ˆ 21:4, JFF ˆ 11:0 Hz, 3F), À102.7 ꢀm,
3
3
4
1F), À155.8 ꢀddq, JFF ˆ 133:8, JFH ˆ 24:4, JFF
ˆ
3
3
21:6 Hz, 1F), À172.7 ꢀdqd, JFF ˆ 133:2, JFF ˆ 10:7,
4JFH ˆ 4:4 Hz, 1F) ppm. {1H} 19F NMR ꢀCDCl3): À68.6
4
3
ꢀdd, JFF ˆ 21:4, JFF ˆ 11:0, 3F), À102.9 ꢀs, 1F), 155.9
4.3.6. 1-+3-Fluorophenyl)-+2E)-2,3,4,4,4-but-en-1-ol
+6), +E)-CF3CF=CFCH+OH)-3-FC6H4
3
4
ꢀdq, JFF ˆ 133:9, JFF ˆ 21:6 Hz, 1F), À172.8 ꢀdq,
3JFF ˆ 133:2, 3JFF ˆ 10:7 Hz, 1F) ppm. 13C NMR ꢀCDCl3):
66.4 ꢀd, 2JCF ˆ 21:0 Hz), 127.5±130.9 ꢀm), 130.3±155.6 ꢀm)
Similarly, 1.07 g ꢀ16.5 mmol) Zn/Ag couple, 5.0 ml dry
DMF, 0.66 g ꢀ5.3 mmol) of 3-¯uorobenzaldehyde and
2.71 g ꢀ10.5 mmol) of ꢀZ)-CF3CF=CFI gave 0.65 g ꢀ48%)
of a 70:30 mixture of ꢀ6) and 3-¯uorobenzaldehyde. 1H
‡
ppm. GC±MS: 256 ꢀM , 11), 239 ꢀ21), 236 ꢀ10), 125 ꢀ100),
123 ꢀ49), 97 ꢀ60). HRMS: calcd. for C10H6F6O: 256.0323;
found: 256.0324.
A similar Barbier reaction using Zn/Ag couple gave 51%
ꢀ3) as a 52:48 mixture of ꢀ3) and 4-¯uorobenzaldehyde.
3
NMR ꢀCDCl3): 2.9 ꢀbrs, 1H), 5.7 ꢀdd, JFH ˆ 24:3,
4JFH ˆ 4:3 Hz, 1H), 6.9±7.6 ꢀm, 4H) ppm. 19F NMR
4
3
ꢀCDCl3): À68.6 ꢀdd, JFF ˆ 21:8, JFF ˆ 10:8 Hz, 3F),
3
3
À112.1 ꢀm, 1F), À155.9 ꢀddq, JFF ˆ 134:0, JFH ˆ 23:5,
3
3
4.3.4. 1-+4-Trifluoromethylphenyl)-+2E)-2,3,4,4,4-
pentafluorobut-en-ol +4), +E)-CF3CF=CFCH+OH)-
4-CF3C6H4
4JFF ˆ 21:4 Hz, 1F), À171.8 ꢀdqd, JFF ˆ 134:0, JFF
ˆ
10:9, JFH ˆ 4:6 Hz, 1F) ppm. {1H} 19F NMR ꢀCDCl3):
4
4
3
À68.6 ꢀdd, JFF ˆ 21:9, JFF ˆ 10:8 Hz, 3F), À112.1 ꢀs,
1F), À155.9 ꢀdq 3JFF ˆ 134:2, 4JFF ˆ 21:3 Hz, 1F), À171.8
Similarly, 0.97 g ꢀ15 mmol) Zn/Ag couple, 5.0 ml dry
DMF, 0.88 g ꢀ5.1 mmol) 4-tri¯uoromethylbenzaldehyde
and 2.63 g ꢀ10.2 mmol) of ꢀZ)-CF3CF=CFI gave 0.86 g
ꢀ55%) of ꢀ4) ꢀ90% purity). 1H NMR ꢀCDCl3): 3.2 ꢀbs,
ꢀdq, JFF ˆ 134:3, JFF ˆ 10:8 Hz, 1F) ppm. 13C NMR
3
3
2
3
ꢀCDCl3): 66.7 ꢀdd, JCF ˆ 21:9, JCF ˆ 1:2 Hz), 113.4
3
2
ꢀd, JCF ˆ 23:4 Hz), 116.0 ꢀd, JCF ˆ 23:4 Hz), 121.7
3
4
4
3
1H), 6.0 ꢀdd, JFH ˆ 23:8, JFH ˆ 3:3 Hz, 1H), 7.2±7.9
ꢀd, JCF ˆ 2:9 Hz), 126.1 ꢀd, JCF ˆ 2:8 Hz), 130.5
3 1
ꢀm, 4H) ppm. 19F NMR ꢀCDCl3): À63.4 ꢀs, 3F), À68.7
ꢀd, JCF ˆ 8:1 Hz), 163.0 ꢀd, JCF ˆ 247:7 Hz), 115±155
‡
ꢀdd, 4JFF ˆ 21:4, 3JFF ˆ 10:8 Hz, 3F), À155.9 ꢀddq, 3JFF
ˆ
ꢀm) ppm. GC±MS: 256, ꢀM , 28), 239 ꢀ37), 125 ꢀ95), 97
ꢀ100). HRMS: calcd. for C10H6F6O: 256.0323; found:
256.0326.
3
4
134:2, JFH ˆ 24:4, JFF ˆ 21:2 Hz, 1F), À171.9 ꢀdqd,
3
4
3JFF ˆ 134:3, JFF ˆ 10:7, JFH ˆ 4:3 Hz, 1F) ppm. {1H}
19F NMR ꢀCDCl3): À63.4 ꢀs, 3F), À68.7 ꢀdd, JFF ˆ 21:4,
4
3JFF ˆ 10:8 Hz, 3F), À156.1 ꢀdq, JFF ˆ 134:2, JFF
ˆ
4.3.7. 1-+4-Chlorophenyl)-+2E)-2,3,4,4,4-pentafluorobut-
en-1-ol+ 7), +E)-CF3CF=CFCH+OH)-4-ClC6H4
3
4
3
3
21:2 Hz, 1F), À171.9 ꢀdq, JFF ˆ 134:3, JFF ˆ 10:7 Hz,
1F) ppm. 13C NMR ꢀCDCl3): 66.5 ꢀd, JCF ˆ 21:7 Hz),
Similarly, 0.97 g ꢀ15 mmol) Zn/Ag couple, 5.0 ml dry
DMF, 0.71 g ꢀ5.1 mmol) of 4-chlorobenzaldehyde and
2.61 g ꢀ10.1 mmol) of ꢀZ)-CF3CF=CFI gave 0.73 g ꢀ53%)
2
‡
120±125.9 ꢀm), 130.3±175.6 ꢀm) ppm. GC±MS: 306 ꢀM ,
35), 287 ꢀ46), 286 ꢀ10), 237 ꢀ83), 217 ꢀ100), 173 ꢀ64), 159
ꢀ65), 145 ꢀ66). HRMS: calcd. for C11H6F8O: 306.0291;
found: 306.0291.
1
of a 74:26 mixture of ꢀ7) and 4-chlorobenzaldehyde. H
3
4
NMR ꢀCDCl3): 4.1 ꢀbs, 1H), 5.6 ꢀdd, JFH ˆ 24:7, JFH
ˆ
4:5 Hz, 1H), 6.8±7.2 ꢀm, 4H) ppm. 19F NMR ꢀCDCl3):
4
3
4.3.5. 1-+2-Trifluoromethylphenyl)-+2E)-2,3,4,4,4-
pentafluorobut-en-1-ol +5), +E)-CF3CF=CFCH+OH)-
2-CF3C6H4
À68.6 ꢀdd, JFF ˆ 21:9, JFF ˆ 10:8 Hz, 3F), À155.8
3 3 4
ꢀddq, JFF ˆ 134:3, JFH ˆ 24:2, JFF ˆ 21:8 Hz, 1F),
À172.3 ꢀdqd, 3JFF ˆ134:3, 3JFF ˆ 10:7, 4JFH ˆ 4:4 Hz, 1F)
4
Similarly, 0.99 g ꢀ15 mmol) Zn/Ag couple, 5.0 ml dry
DMF, 0.85 g ꢀ4.2 mmol) 2-tri¯uoromethylbenzaldehyde
and 2.54 g ꢀ9.85 mmol) of ꢀZ)-CF3CF=CFI gave 0.67 g
ꢀ52%) of a 60:40 mixture of ꢀ5) and 2-tri¯uoromethylben-
zaldehyde. 1H NMR ꢀCDCl3): 4.1 ꢀbrs, 1H), 6.0
ppm. {1H} 19F NMR ꢀCDCl3): À68.6 ꢀdd, JFF ˆ 21:9,
3
4
3JFF ˆ 10:9, 3F), À155.8 ꢀdq, JFF ˆ 134:0, JFF
ˆ
3
3
21:2 Hz, 1F), À172.3 ꢀdq, JFF ˆ 134:3, JFF ˆ 10:7 Hz,
1F) ppm. 13C NMR ꢀCDCl3): 66.4 ꢀd, 2JCF ˆ 21:4 Hz), 111±
116 ꢀm), 127±132 ꢀm), 161±171 ꢀm), ppm. GC±MS: 272
‡
3
4
ꢀdd, JFH ˆ 22:4, JFH ˆ 3:8 Hz, 1H), 7.2±7.8 ꢀm, 4H)
ꢀM , 34), 237 ꢀ60) 217 ꢀ50), 139 ꢀ100), 131 ꢀ10), 111 ꢀ75).
ppm. 19F NMR ꢀCDCl3): À59.7 ꢀd, JFH ˆ 6:9 Hz, 3F),
HRMS: calcd. for C10H6F5O35Cl: 272.0027; found:
272.0021.
4
4
3
À68.9 ꢀdd, JFF ˆ 21:6, JFF ˆ 10:7 Hz, 3F), À155.0
3
3
4
ꢀddq, JFF ˆ 133:3, JFH ˆ 22:4, JFF ˆ 21:7 Hz, 1F),
A similar reaction with acid-washed zinc gave 51% of a
similar product mixture.
3
3
4
À169.3 ꢀdqdq, JFF ˆ 133:3, JFF ˆ 10:6, JFH ˆ 3:8,

V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
239
4.3.8. Attempted preparation of 1-+4-methyoxyphenyl)-
+2E)-2,3,4,4,4-but-en-1-ol+ 8), +E)-CF3CF=CFCH+OH)-
4-MeOC6H4
4.3.13. 1-Phenyl-+2E)-2,3,4,4,4-pentafluorobut-en-1-ol
+1) via reaction of benzaldehyde with pregenerated
+Z)-CF3CF=CFZnI
Similarly 0.65 ꢀ10 mmol) of acid-washed zinc, 5 ml dry
DMF, 0.27 g ꢀ2.0 mmol) of 4-anisaldehyde and 1.0 g
ꢀ4.0 mmol) of ꢀZ)-CF3CF=CFI gave 0% ꢀ8). 19F NMR
analysis of the reaction mixture indicated only ꢀZ)-
CF3CF=CFZnI ꢀspectroscopic data in agreement with the
zinc reagent generated from ꢀZ)-CF3CF=CFI and Zn). There
was no trace of ꢀ8) detected in the reaction mixture by 19F
NMR analysis.
Similarly, 0.38 g ꢀ3.6 mmol) of benzaldehyde was reacted
with 2.8 mmol of pregenerated ꢀZ)-CF3CF=CFZnI. After 3 h
at room temperature, 19F NMR analysis of the reaction
mixture indicated only the zinc reagent and 0% ꢀ1). The
reaction mixture was then heated to 508C ꢀ3 h), 608C ꢀover-
night) and 808C ꢀ3 h); 19F NMR analysis indicated 0% ꢀ1) at
any of these temperatures.
4.3.14. Effect of 1,4-dinitrobenzene on the reaction of
+Z)-CF3CF=CFI with 2-fluorobenzaldehyde and
Zn/Ag couple
4.3.9. Attempted preparation of 1-+4-methylphenyl)-
+2E)-2,3,4,4,4-pentafluorobut-en-1-ol +9)
Similarly, 0.66 g ꢀ10.0 mmol) Zn/Ag couple, 5 ml dry
DMF, 0.24 g ꢀ2.0 mmol) of 4-tolualdehyde and 1.2 g
ꢀ4.7 mmol) of ꢀZ)-CF3CF=CFI gave 0% ꢀ9). 19F NMR
analysis of the reaction mixture indicated only ꢀZ)-
CF3CF=CFZnI ꢀspectroscopic data in agreement with the
zinc reagent generated from ꢀZ)-CF3CF=CFI and Zn) and
showed no trace of ꢀ9).
Similarly, to 0.36 g ꢀ5.5 mmol) acid-washed zinc, 0.51 g
ꢀ3.0 mmol) of 1,4-dinitrobenzene, 3.6 ml dry DMF, and
0.24 g ꢀ1.9 mmol) of 2-¯uorobenzaldehyde were added
0.96 g ꢀ3.7 mmol) of ꢀZ)-CF3CF=CFI. No reaction ensued
after 1.5 h at RT or heating at elevated temperature. No zinc
reagent was formed and no ꢀ2) was formed.
4.3.15. Pregeneration of +Z)-CF3CF=CFZnI
4.3.10. Attempted preparation of 2,2-dimethyl-+4E)-
4,5,6,6,6-pentafluorohexen-3-ol +10),
+E)-CF3CF=CFCH+OH)C+CH3)3
Similarly, 0.65 g ꢀ10 mmol) of acid-washed zinc, 5.0 ml
dry DMF, 0.18 g ꢀ2.1 mmol) of pivaladehyde and 1.0 g
ꢀ4.0 mmol) of ꢀZ)-CF3CF=CFI gave 0% of ꢀ10). 19F
NMR analysis of the reaction mixture indicated only ꢀZ)-
CF3CF=CFZnI ꢀspectroscopic data in agreement with the
zinc reagent generated from ꢀZ)-CF3CF=CFI and Zn) and
showed no trace of ꢀ10).
A 25 ml ¯ask equipped with a side arm, stir bar, cold
water condenser and attached to a N2 tee was charged with
0.66 g ꢀ10 mmol) of acid-washed zinc and 5.0 ml dry DMF.
To this slurry was added 1.30 g ꢀ5.03 mmol) of ꢀZ)-
CF3CF=CFI. Reaction initiated after a short induction per-
iod as indicated by a exotherm. After cooling to room
temperature and stirring for 2 h, the solids were allowed
to settle and an 19F NMR spectrum was taken of the super-
4
natant liquid. 19F NMR ꢀCDCl3): À67.6 ꢀdd, JFF ˆ 22:6,
3
4
3JFF ˆ 13:1 Hz, 3F), À148.5 ꢀdq, JFF ˆ 108:1, JFF
ˆ
3
3
21:6 Hz, 1F), À179.8 ꢀdq, JFF ˆ 111:3, JFF ˆ 12:1 Hz,
1F) ppm; consistent with the literature data [23]. The yield
of the zinc reagent varied from 69 to 76% in several trials.
4.3.11. 1-+2-Fluorophenyl)-+2Z)-2,3,4,4,4-
pentafluorobut-en-1-ol +11), +Z)-CF3CF=CFCH+OH)-
2-F C6H4
Similarly, 0.88 g ꢀ14 mmol) Zn/Ag couple, 5.0 ml dry
DMF, 0.5 g ꢀ4 mmol) of 2-¯uorobenzaldehyde and 2.1 g
ꢀ8.0 mmol) of ꢀE)-CF3CF=CFI gave 0.51 g ꢀ50%) of a 80:20
mixture of ꢀ11) and 2-¯uorobenzaldehyde. 1H NMR
Acknowledgements
We thank the National Science Foundation for ®nancial
support of this work.
3
ꢀCDCl3): 4.2 ꢀbrs, 1H), 6.0 ꢀd, JFH ˆ 27:8 Hz), 6.7±7.7
ꢀm, 4H) ppm. 19F NMR ꢀCDCl3): À73.6 ꢀm, 3F), À125.1 ꢀd,
3JFF ˆ 4:6 Hz, 1F), À145.6 ꢀm, 1F), À162.7 ꢀq,
3JFF ˆ 12:2 Hz, 1F) ppm. 13C NMR ꢀCDCl3): 62.0 ꢀdd,
References
3
2JCF ˆ 21:6, JCF ˆ 4:5 Hz), 115.6±130.8 ꢀm), 110.1±
‡
154.7 ꢀm) ppm. GC±MS: 256 ꢀM , 60), 236 ꢀ40), 208
ꢀ32), 169 ꢀ65), 123 ꢀ100). HRMS: calcd. for C10H6F6O:
256.0323; found: 256.0329.
[1] C. Blomberg, Barbier Reaction and Related One-Step Processes,
Springer, Berlin, 1993 and references therein.
[2] P. Barbier, Compt. Rend. 128 ꢀ1899) 110.
[3] E. Frankland, J. Chem. Soc. 2 ꢀ1849) 263.
[4] E. Frankland, Leibigs Ann. Chem. 71 ꢀ1849) 171.
[5] H. Blancoou, P. Morreau, A. Commeyras, J. Chem. Soc., Chem.
Commun. ꢀ1976) 885.
4.3.12. +2E)-1,1,1,2,3-pentafluorohepten-4-ol +12),
+E)-CF3CF=CFCH+OH)CH2CH2CH3
[6] C. Wakselman, M. Tordeux, Bull Soc. Chim. France ꢀ1986) 868.
[7] T. Kitazume, N. Ishikawa, Chem. Lett. ꢀ1981) 1679.
[8] T. Kitazume, T. Ikeya, J. Org. Chem. 53 ꢀ1988) 2349±2350.
[9] M. Fujita, T. Hiyama, K. Kondo, Tetrahedron Lett. 27 ꢀ1986)
1239.
Similarly, 0.71 g ꢀ11 mmol) Zn/Ag couple, 5.0 ml dry
DMF, 0.31 g ꢀ4.3 mmol) of butanal and 2.2 g ꢀ8.4 mmol)
of ꢀZ)-CF3CF=CFI gave 0% ꢀ12). 19F NMR analysis of
the reaction mixture indicated only ꢀE)-CF3CF=CFH and
ꢀZ)-CF3CF=CFZnI in a 1:1.4 ratio.
[10] N.J. O'Reilly, M. Maruta, N. Ishikawa, Chem. Lett. ꢀl984) 517.

240
V. Jairaj, D.J. Burton / Journalof Ful orine Chemistry 111 +2001) 233±240
[11] C. Francese, M. Tordeux, C. Wakselman, J. Chem. Soc., Chem.
Commun. ꢀ1987) 642.
[19] S.W. Hansen, T.D. Spawn, D.J. Burton, J. Fluorine Chem. 35 ꢀ1987)
219.
[12] C. Francese, M. Tordeux, C. Wakselman, Tetrahedron Lett. 29 ꢀ1988)
1029.
[20] T.D. Spawn, Ph.D. Thesis, University of Iowa, 1987.
[21] T. Ishihara, K. Yamaguchi, M. Kuroboshi, Chem. Lett. ꢀ1989) 1191.
[22] T. Ishihara, K. Yamaguchi, M. Kuroboshi, Y. Okada, J. Org. Chem.
55 ꢀ1990) 3107.
[13] Y. Shen, M. Qi, J. Fluorine Chem. 66 ꢀ1994) 175.
[14] R. Meithchen, C. Zur, D. Peters, Synthesis ꢀ1998) 1033.
[15] C. Hu, X. Tang, F. Hong, J. Chem. Soc., Chem. Commun. ꢀ1994)
289.
[23] P.L. Heinze, D.J. Burton, J. Org. Chem. 53 ꢀ1988) 2714.
[24] J.M. Conia, G. Rousseau, Tetrahedron Lett. 22 ꢀ1981) 649.
[25] P.L. Heinze, T.D. Spawn, D.J. Burton, S. Shin-Ya, J. Fluorine Chem.
38 ꢀ1988) 131.
[16] C. Blomberg, Chem. Ind. ꢀHandbook of Grignard Reagents) 219.
[17] C. Blomberg, Chem. Ind. ꢀHandbook of Grignard Reagents) 249.
[18] K.J. Klabunde, M.S. Key, J.Y.F. Low, J. Am. Chem. Soc. 94
ꢀ1972) 999.
[26] D.J. Burton, T.D. Spawn, P.L. Heinze, A.R. Bailey, S. Shin-Ya, J.
Fluorine Chem. 44 ꢀ1989) 167.